My Goals

The goal of this experiment is to see if there is a link between the phenotype of a newt or snake and its ability to survive. This prediction stems from the idea that if a phenotype is higher in a newt or snake, the individual will survive an interaction with the opposite species. To look at this I will calculate the correlation between an individuals phenotype and an individuals SLiM fitnessScaling (I am calling it beta). There are some potential problems with this calculation, like space (some areas one phenotype might be beneficial, but it others not) and that the interaction is not very strong.

Beta calculation for the Newt: beta_n = cor(phenotypes_n, inds_n.fitnessScaling); Beta calculation for the Snake: beta_s = cor(phenotypes_s, inds_s.fitnessScaling); Must be calculated in early. fitnessScaling set to 0 if newt or snake is killed by the opposite species, but what about interspecific interactions (density dependence)

First, I will look to see if there is an interaction between the newt and snake phenotypes and the amount of newts and snakes that are being killed (specifically killed by the other species). I predict that when snakes have a higher phenotype more newts will be killed and less snakes will be killed. When the phenotype of the newt is higher I predict that more snakes will be killed. When the phenotypes of newts and snakes are similar I predict that both newts and snakes will be killed at the same rate. Previously, I had seen that there were less newts being killed when the snake phenotype was higher.

I thought the results were strange (seems like less newts dying when snake phenotype is higher) so I looked into further into what might be happening. I recorded the number of newt and snake deaths (cause by the other species) every 20 generations and recorded the population sizes for each species.

I found that there was an interaction between the difference in phenotype of newts and snakes and survivability of sorts. When the phenotype of the snake is higher less snakes were dying, but so where less newts. Less newts dying is probably due to the population size of newts being diminished when the phenotype of the snake is high. It’s possible that the most resistant snakes ate all of the newts in that area and just had fewer newts to eat. The reverse also happens when newts have a higher phenotype. There are less snakes and newts dying, bit the population size of snakes is small. Red and blue xmas tree and X marks the spot.

Sinse there seems to be an interaction between phenotype and population size, I will look at how the population size lines up with the mean phenotype of newts and snakes. I predict that there will be an arc of color ranging from light to dark depending on which species has a higher population size.

I made some nice whale graphs! When the newt population size is large and the snakes population size is small the newt phenotype is high. When the newt population size is small and the snake population size is large the newts have a lower phenotype. Something strange happens with the phenotype of the snake, it is higher when the newt population size is large and the snake population size is small. The snake phenotype is also high when the snake population size is large and the newt population size is small. My theory is that the few remaining snakes in an area of very toxic newts are very resistant, but why don’t we see the opposite happening in newts? <- this also happens in the snake-newt papers weird. I would be wary of the whale graphs because these are a combined results of 107 simulations with a sample collected every 20 generations. It might be different in each separate simulation.

Next, I look into the betas and am not sure what I expect. I know the interaction between newts and snakes is small each generation, but there seems to be some interaction between phenotype and the ability for an individual to survive.

Lets look at the distribution

The beta values for newts and snakes seem to be similar. The means are also very close.

Lets plot the beta values for the newt against the beta values for the snake.

I present to you the green eyeball. It looks like there is no correlation between newt beta values and snake beta values. Prediction: this might be due to space, some areas might have correlation while other don’t. This is one value for the entire area

Finally, the last few graphs are extra things to look at.

The interaction of Newts and Snakes

I was wondering how often newts and snakes interact in a SLiMulation so I had SLiM print out four values, snake_found_newt (when a snake was potentially able to find a newt), newt_found (a snake found a newt), snake_deaths (snakes that were killed by newts), and newt_deaths (newts that were eaten by snakes). I also added the population size for newts and snakes to see in the number of interactions depend on the population size - which is linked to density. It looks like one the population size stabilizes (around 100 generations) there are 30-60 interactions each generation. Today 8/30/2021 I will run 20 of the same simulation to see if I can get a consensus on how many times newts and snake interact in each generation. Once I get this value I can plug it into nospace.slim to compare the results of the same simulation to a simulation with a lack of space. This will hopefully help me understand how space effects the coevolution of newts and snakes

It seems like newts and snakes interact between 20 and 60 times each generation. The mean is around 40. Snakes also seem to have a higher population size. The number of interactions become steady around 150 generations. As the number of interactions increase the number of death increase and its even between both species

Now, I am just going to run some numbers to see what the interaction rate was in my 1on1 slim simulation and what is should be set to in my nospace slim. In my 1on1 slimulations I us an interaction rate of 0.05 and surprisingly when I reverse calculate it from the mean number of interactions and mean population size I get 0.05 so it seems like it is working which is good (population specific to snake pop size).

## [1] "If the population size is 700 and the number of interactions is 40, then the interation rate is 0.0571428571428571."
## [1] "If the population size is 600 and the number of interactions is 40, then the interation rate is 0.0666666666666667."
## [1] "If the population size is 800 and the number of interactions is 40, then the interation rate is 0.05."
## [1] "Now if I pick the interaction rate of 0.05 and the number of individuals is 700, the number of interactions should be 35"

This light green cloud shows that when the population size is larger there are more interactions. This picture reminds me of a rain cloud.

It seems like using the formula population_size*interaction_rate will get me the correct number of interactions every time so I should devise a way to have a normal distribution with a mean around 40.

No Space SLiM Results

next one